SHAKE TABLE TEST ON 3-STOREY LIGHT-FRAME TIMBER BUILDING

Size: px

Start display at page:

Download "SHAKE TABLE TEST ON 3-STOREY LIGHT-FRAME TIMBER BUILDING"

Maximilian Tyrone Doyle
6 years ago
Views:

1 SHAKE TABLE TEST ON 3-STOREY LIGHT-FRAME TIMBER BUILDING Tiziano Sartori 1, Daniele Casagrande 2, Roberto Tomasi 3, Maurizio Piazza 4 ABSTRACT: Shake table tests were carried out on a 7m x 5m three stories timber light frame building (7.5 m high) at the TreesLab laboratory (Eucentre) in Pavia. The goal of the research project was to evaluate the seismic behavior of a typical Italian prefabricated timber building and to study the interaction between the singular structural components tested in quasi-static way. The 1979 Montenegro Earthquake ground motion, recorded at Ulcinj-Hotel Albatros, was selected as the ground motion for seismic tests. The maximum peak ground acceleration was scaled to 0.07g, 0.27g, 0.5g. 0,7g and 1g in order to evaluate the performance of the building at different levels of seismic inputs. A frequency and damping evaluation tests were used before and after each seismic test to identify natural frequencies, modal shapes and equivalent viscous damping ratio, exciting the structure with a low amplitude white random noise. After each seismic excitations the specimen was inspected for evidences of damage. The building designed for a 0,28g PGA representing a hazard level of 10% probability of exceedance in 50 years or a return period of 475 years, showed no visual damages at all the stage of the tests. KEYWORDS: wood-frame multi-storey buildings, seismic performance, shake table test 1 INTRODUCTION TIMBER-FRAMED MULTI-STOREY BUILDINGS The timber-frame multi-storey buildings are gaining increasing importance, presenting a sustainable alternative to solutions made with other building materials, as happened during the reconstruction following the earthquake in Abruzzo in 2009 (CASE project). In fact, as demonstrated by the behaviour of wooden buildings in areas of high seismic risk, the low density of the material has as an advantage the minor stresses induced by a seismic event. On the other hand, some systems such as wood-framed buildings (the most popular for civil construction are in the United States, Scandinavia, Canada and New Zealand) are characterized by great structural redundancy of connections with good mechanical capacity of ductility and energy dissipation, for which then, in accordance 1 Tiziano Sartori, Department of mechanical and structural Trento, Italy. tiziano.sartori@ing.unitn.it 2 Daniele Casagrande, Department of mechanical and structural Trento, Italy. daniele.casagrande@ing.unitn.it 3 Roberto Tomasi, Department of mechanical and structural Trento, Italy. roberto.tomasi@ing.unitn.it 4 Maurizio Piazza, Department of mechanical and structural Trento, Italy. maurizio.piazza@ing.unitn.it with the recent regulations (Eurocode 8, NTC 2008), is assigned one of the highest q-factor. The frame construction system is emerging in our country, thanks to the high level of prefabrication, but with some peculiarities, both in terms of geometry and materials used, compared to the lightweight solutions typically adopted in the North American and North European market: the vertical stud and transoms have a major sections and also the cladding panels are preferably thicker and heavier. 1.2 INDUSTRIAL RESEARCH PROJECT CHI QUADRATO For these previous mentioned reasons it has been considered necessary as part of an industrial research project dedicated to sustainable wooden buildings, carry an extensive experimental campaign having as its goal the mechanical characterization in terms of stiffness, strength and ductility of components and structural links of a wooden framed construction system when the geometric parameters and materials used, vary. As part of this research angular connection between the wall frame and horizontal partitions, connections between the top plates and wooden covering sheets and horizontal wall framed have been subjected to static tests in accordance with the protocols of the rules of monotonic and cyclic type [1,2].

2 DESIGN AND DESCRIPTION OF THE BUILDING 2.

structural panels The horizontal elements are formed by box section elements 140 mm height to which are superimposed and nailed with OSB panels with 15 mm thickness in order to ensure the behavior at

Figure 2: Test Set-up of the tests on framed walls subject to horizontal action (monotonic and cyclic) and different magnitude of vertical load The results of this experimental campaign have allowed

through linear (modal response spectrum analysis) and non-linear methods (analysis of push-over, time-history analysis). 1.

2 2 DESIGN AND DESCRIPTION OF THE BUILDING 2.1 GEOMETRY OF THE BUILDING AND THE STRUCTURAL ELEMENTS The test building has a rectangular plan, 5m x 7m and it is on three levels (ground floor, first floor and attic) to a maximum height to the peak of 7.65 m. Figure 1: Test Set-up of the tests on the connections between vertical stud and sheating panel and the connections between wall and foundation Figure 3: Three-dimensional view of the building without structural panels The horizontal elements are formed by box section elements 140 mm height to which are superimposed and nailed with OSB panels with 15 mm thickness in order to ensure the behavior at rigid in-plane diaphragm. Figure 2: Test Set-up of the tests on framed walls subject to horizontal action (monotonic and cyclic) and different magnitude of vertical load The results of this experimental campaign have allowed the optimization of the manufacturing technology and the creation of an experimental database which has been used in the analysis of the response of multi-storey buildings in seismic areas, both through linear (modal response spectrum analysis) and non-linear methods (analysis of push-over, time-history analysis). 1.3 SCOPE OF TESTS In order to validate the algorithms based on the knowledge of basic structural components, it has been designed and built a multi-storey wooden building in full scale to be tested on a shake table. Important it is the comparison between the current calculation methods and assumptions implemented in the Regulations and the results of experiments. It therefore seems essential to determine which is the behavior of individual structural components within the building in full scale subjected to seismic action, how they interact and what are the possible failure mechanisms. Figure 4: Testing building The roof with two pitches has been made by solid wood beams and a wooden plank with stiffened perforated metal strips. The structural walls are framed with 15 mm thickness OSB panels, fixed both sides by ring nails 2.8 x 60 mm. Uprights and transoms of the wall has been made with solid wood, section 160 x 60 and 160 x 100 mm. The building has been tested without finish and nonstructural components. In order to simulate the additional mass relative to the permanent loads and accidental loads provided in case of seismic event, they have been placed on the floors some concrete blocks of equivalent weight. In order to simulate the mass of the insulation and the entire stratigraphy of the roof, they have been placed on

the roof itself the tiles weighed and additional materials within the walls.

Table 1: Seismic hazard parameters in relation to the different limit condition Figure 5: Additional masses on the floors and inside the structural wall 2.

(Figure 6b). The tensile forces in the corners of the walls have been assigned to hold-down, to the foundation plan (Figure 6c) and to metal plates nailed to the upper floors (Figure 6d).

LIMIT T R a g F 0 T C CONDITION SLO 30 0,049 2,418 0,249 SLD 50 0,068 2,518 0,268 SLV 475 0,277 2,280 0,424 SLC 975 0,402 2,328 0,476 2.

been developed. In particular, they have been used some linear models in order to evaluate the dynamic parameters of the building, useful for a correct choice of the input parameters.

3 the roof itself the tiles weighed and additional materials within the walls. forces of the project so calculated have been divided between the walls in proportion to their structural rigidity, assuming rigid floors. Table 1: Seismic hazard parameters in relation to the different limit condition Figure 5: Additional masses on the floors and inside the structural wall 2.2 GEOMETRY AND ORGANIZATION OF THE CONNECTION SYSTEMS In order to ensure the interaction between the various structural elements and the connection of the building with the foundation, they have been used different connecting devices. For the horizontal sliding of the structural walls, they have been used plates nailed to the upper floors (Figure 6a) and pairs of wood screws 8x180 inclined at 45 to the plane of the foundation (Figure 6b). The tensile forces in the corners of the walls have been assigned to hold-down, to the foundation plan (Figure 6c) and to metal plates nailed to the upper floors (Figure 6d). The connection of the floor with the walls has been carried out by wood screws 8x220. LIMIT T R a g F 0 T C CONDITION SLO 30 0,049 2,418 0,249 SLD 50 0,068 2,518 0,268 SLV 475 0,277 2,280 0,424 SLC 975 0,402 2,328 0, ESTIMATION OF PARAMETERS OF THE TEST BY PRELIMINARY NUMERICAL ANALYSIS In order to estimate the parameters necessary for the design of the shake table test, some numerical simplified models have been developed. In particular, they have been used some linear models in order to evaluate the dynamic parameters of the building, useful for a correct choice of the input parameters. a) b) c) d) Figure 6: a: Nail plates used to prevent horizontal translation of floors - b: Screws inserted at 45 used to prevent the horizontal translation of walls at foundation level - c: Hold-down - d: Tie-down Figure 7: Numerical model Each wall was modelled by an equivalent diagonal spring, calibrated in two different ways. In the first one (mod. A) the OSB shear deformation contribution was implemented; in the second one (mod. B) the sheeting nail-slip stiffness contribution was added the OSB shear deformation. The hold-down and angle brackets contribution was neglected since the frequency evaluation tests would have been under a low intensity exciting. The predicted longitudinal fundamental period was equal to 0.11 s for mod. A and to 0.30 s for mod. B. 2.3 STRUCTURAL PROJECT The design has been made in accordance with Italian Standard (NTC2008) integrated with the European provisions Eurocode 5 and Eurocode 8. The building is assumed to be a residential house located in site characterized by the maximum value of seismic hazard with a return period of 475 years, as prescribed by Italian law. The seismic design has been carried out by static analysis assuming a behaviour factor q equal to 4. The

a) b) Figure 8: Elastic mode shapes A non-linear model has been implemented for the estimation of the base shear

These analyses have also made it possible to design the test phases at different intensities of seismic input, in

1 ASSEMBLY The assembly of the building has been done directly on the vibrating bench.

system and the high accuracy of the machining in factory permitted to complete the phases of assembly in 3 days by 5

During the assembly, in fact, it was necessary to perform only the placement of structural elements (walls, floor

paragraph 2.2. In Figure 10, they are visible the different phases of assembly. g) h) Figure 10: Assembly phases.

and placement e: Second floor walls placement f: Third floor walls placement o g: Roof beams placement h: Roof

2 SET-UP OF THE MEASUREMENTS The measures concerned accelerations, displacements and deformations.

acquisition system which can monitor in continuous the displacement of the markers placed on the building, thanks to the

4 a) b) Figure 8: Elastic mode shapes A non-linear model has been implemented for the estimation of the base shear capacity and of the over turning moment (OTM). These analyses have also made it possible to design the test phases at different intensities of seismic input, in relation to the expected limit states. c) d) e) f) 3 SEISMIC TESTS ON SHAKE TABLE 3.1 ASSEMBLY The assembly of the building has been done directly on the vibrating bench. The connection between the building and the table has been realized by a grid of steel beams arranged in correspondence of the structural walls. Figure 9: Grid of steel beams arranged in correspondence of the structural walls The significant prefabrication of the system and the high accuracy of the machining in factory permitted to complete the phases of assembly in 3 days by 5 workers. During the assembly, in fact, it was necessary to perform only the placement of structural elements (walls, floor panels, stiffened panels, primary and secondary warping of the roof) connecting them with the systems described in paragraph 2.2. In Figure 10, they are visible the different phases of assembly. g) h) Figure 10: Assembly phases. a: Structural elements carriage - b: First floor walls placement c: 1st Floor elements placement d: OSB boards nailing and placement e: Second floor walls placement f: Third floor walls placement o g: Roof beams placement h: Roof Perforated metal strips 3.2 SET-UP OF THE MEASUREMENTS The measures concerned accelerations, displacements and deformations. Data were acquired with a sampling frequency of 1024 Hz: 103 instruments have been positioned in addition to the optical acquisition system which can monitor in continuous the displacement of the markers placed on the building, thanks to the various infrared cameras (sampling frequency: 60 Hz). The positioning of the measurement s instruments has taken place taking into account the mono directionality of motion of the bench, therefore on the walls parallel to the seismic and structural elements most stressed. The following measurement s instruments have been installed for all 12 monitored walls (6 on the first floor and 6 on the second) : displacement transducer capable of measuring horizontal scrolling, the vertical displacement and the shear deformation of the walls; strain capable of measuring the deformation of the metal elements of connection (hold-down and tiedown).

The following Table 2 recaps the measurement s instruments installed. plane and the relative sliding of the base docks compared to the steel base.

28 Wall upflit and slippage Accelerometers 27 Accelerations It was decided to not monitor the walls of the attic as not subject to significant shifts.

points of the walls of the front C (Figure 11). a) b) Figure 12: Walls monitored First floor 3.

"Ulcinj - Hotel Albatros", located at an epicentral distance of 21 km, during the Montenegro earthquake of 15/04/1979 (Mw 6.9). e) f) Figure 11: Instrumentation.

for in plane deformation of horizontal diaphragms e: Interior wall inter-storey drift potentiometer f: Optical acquisition system markers The acceleration measurements have

8 accelerometers on the eaves and 2 on the peak have been adopted on the roof. The acceleration measurements have been done in the horizontal directions, only.

5 The following Table 2 recaps the measurement s instruments installed. plane and the relative sliding of the base docks compared to the steel base. Table 2: Measurement s instruments Typology Qty Type of measure Cells hold-down 12 Force Cells tie-down 11 Force Wire 28 Wall shear defomation Potentiometers Potentiometers 28 Wall upflit and slippage Accelerometers 27 Accelerations It was decided to not monitor the walls of the attic as not subject to significant shifts. The use of an optical acquisition system has allowed also to monitor both the absolute displacements of the test building and the relative displacements of the control points of the walls of the front C (Figure 11). a) b) Figure 12: Walls monitored First floor 3.3 TESTING SEQUENCE c) d) The test building has been subjected to 5 different intensity of seismic input, scaling appropriately the accelerogram recorded by the station "Ulcinj - Hotel Albatros", located at an epicentral distance of 21 km, during the Montenegro earthquake of 15/04/1979 (Mw 6.9). e) f) Figure 11: Instrumentation. a: Hold down load cells - b: Wall uplift and slippage potentiometers c: Displacement transducer for horizontal slip between floor elements d: wire displacement transducer for in plane deformation of horizontal diaphragms e: Interior wall inter-storey drift potentiometer f: Optical acquisition system markers The acceleration measurements have been done with a pairs of unidirectional accelerometers disposed on each of the 4 corners of the two intermediate floors. 8 accelerometers on the eaves and 2 on the peak have been adopted on the roof. The acceleration measurements have been done in the horizontal directions, only. Some displacement transducers have been used to monitor the possible deformations of the floors in its Figure 13: Accelerogram. Montenegro earthquake Ulcinj -Hotel Albatros station PGA: 0.224g The choice of this accelerogram occurred in relation to the pseudo-acceleration response spectrum, particularly significant in the range of the theoretical fundamental frequency of the building.

6 4.2 VARIATION OF FUNDAMENTAL PERIOD Figure 14: Pseudo acceeration response spectrum (ξ=5%) The values of maximum acceleration (PGA) at which the building has been tested in the 5 stages of test, are shown in the table below. These values have been derived considering the analysis of numerical simulation as well as to the results obtained after the first test phase. Table 3: Test phases and maximum acceleration PHASE 1 0,07g PHASE 2 0,28g PHASE 3 0,50g PHASE 4 0,70g PHASE 5 1,00g Each of the seismic testing have been preceded by a phase of "tuning", which was necessary to calibrate the control parameters of the shake table in order to obtain a good correspondence between the feedback and the reference. This procedure has been done submitting the building to seismic motion of a low intensity flat white noise. During this phase, we proceeded to the acquisition of the signals of the different accelerometers in order to be able to conduct after the identification procedure of the dynamic structure. Indeed, it is essential to estimate with good accuracy the dynamic properties (fundamental frequencies, vibration modes, damping equivalent) of the test building in order to quantify the damage to the building after each seismic test. The evaluation of the structural damage has also been done after each test by careful visual inspection. It is important to underline that between one phase and following one, any kind of repair on the structure hasn t been done. 4 RESULTS 4.1 INSPECTION RESULTS The inspections performed on the building after all the tests has shown no visible damages to structural elements. The purpose of the frequency evaluation tests was to identify the fundamental period (or frequency) and mode shapes of the test structure after each seismic test. Ten accelerometers time-histories through the structure and one accelerometer on the shake table were analysed to evaluate the transfer functions TFs as ratio, in the frequency domain, between the story acceleration responses and the base motion. The peak amplitude method was used to estimate the fundamental frequency of the longitudinal direction. The data were preprocessed by means of an anti-aliasing filter (frequency of Nyquist equal to 128 Hz), a linear trend removing and the PSD Welch s method computation (using a Hanning window). The sampling frequency was equal to 256 Hz and the frequency resolution to Hz. Figure 15: TF Amplitude and Phase between channel A1 (1 st floor) and base motion after 0.28g test The fundamental period was unchanged until 0.70 g and equal to s (6.125 Hz). After the phase 4 and the phase 5 the period increased respectively to s (6.060 Hz) and to s (5.814 Hz). The fundamental period variation suggests that some damages in the structure happened only for the higher acceleration seismic tests. No significant 1 st mode shape variation was evaluated. Figure 16: Mode shapes before 0.07g test and after 1.00g 4.3 INTER STOREY DRIFT MEASUREMENT One of the most important parameters to relate the input seismic action to the performances of the structure is the inter storey drift measurement, which is can associated a damage level to.

7 The measurements were performed by means of two potentiometers inside the building and processing the data of optical acquisition system outside. The potentiometers data were filtered with a lowpass filter using a passband frequency of 25 Hz and a stopband frequency for 50 Hz with a 40 db attenuation. The optical acquisition system were not filtered. The maximum inter storey drifts for 1 st floor are reported in the following table: Table 4: 1 st floor maximum inter storey drifts 1 st floor Ch 41 (Interior) Optical S. (External) Test [%] 0.07g g g g g HOLD DOWN FORCE MEASUREMENTS The peak tensile hold down forces during each seismic test are reported. No significant forces were registered until the 0.50 g test. The maximum force was measured in the 1.25 m wall hold down an it was equal to 38.8 kn. It s important to underlined that the collapse of hold down was evaluated equal to 50 kn in the monotonic quasi static hold down test. However the increasing of longitudinal fundamental period has highlighted a non-elastic behaviour of the tested structure after 0.70g test. The magnitude of measured 1 st floor inter storey drifts (1.12% for 1.00g test) confirmed this statement according to the results reported in [3] and [4]. The peak tensile hold down measured forces suggest that until 0.50 g test that the role of hold downs could be neglected. For higher input test they become necessary to avoid rocking of the wall. Probably the box structural building behaviour is predominant at lower input tests. 6 FUTURE DEVELOPMENTS The analyses of all instruments are going on in order to investigate the seismic behaviour of the tested structure. Acceleration, Optical system, Tie-Down measures will be examined. Capacity spectra, global hysteretic response and energy response will be also performed. The main goal is to understand the interaction between structural elements under seismic loads. ACKNOWLEDGEMENT The authors gratefully acknowledge the CHI-Quadrato Consortium for partly financing the study, within the research program supported by Autonomous Province of Trento and Legnocase factory to provide the material for the specimens. REFERENCES [1] Conte A., Piazza M., Sartori T., Tomasi R. (2011), Experimental investigation on connections between wood framed shear walls and foundations, Proceedings of the Structural Engineering World Conference, Como Cernobbio, Italy [2] Conte A., Piazza M., Sartori T., Tomasi R. (2011), Influence of sheathing to framing connections on mechanical properties of wood framed shear walls, Congresso Ingegneria Sismica Anidis 2011, Bari [3] CUREE project: [4] NEESWood project: [5] UNI-EN Eurocode 8 Design of structures for earthquanke resistance. Part 1: General rules, seismic actions and rules for building Figure 17: Peak Tensile Hold down forces for 1.00 g [kn] 5 CONCLUSIONS A three-storey light framed timber building shake table tests has been presented. Five seismic tests were performed in order to investigate different structure limit states. The building, designed according to the Italian and European standards, showed high seismic performances: no visible damages were noticed for all tests.

SOFIE Project - Cyclic Tests on Cross-Laminated Wooden Panels

SOFIE Project - Cyclic Tests on Cross-Laminated Wooden Panels Ario Ceccotti Director Marco Pio Lauriola, Mario Pinna, Carmen Sandhaas Researchers CNR-IVALSA Italian National Research Council, Trees and